Computational Biology

Options

Volunteer Basis, Potential for UROP Funding, Potential for Academic Credit

Overview

Title: Deep Learning and Image Analysis for Spatial Transcriptomics in Cancer

Are you interested in cancer biology, bioinformatics, and artificial intelligence? Do you want to learn how deep learning can connect tissue images with gene expression data?

Our lab (www.drieslab.com) at Boston University Medical Campus (BUMC) is seeking a motivated undergraduate student to work on integrating histological imaging with spatial transcriptomics data. Spatial transcriptomics technologies (e.g., Xenium, MERFISH, Visium HD) generate gene expression measurements directly within intact tissue sections and are typically paired with high-resolution H&E-stained images. These images contain rich information about tissue organization, tumor boundaries, immune infiltration, and other structural features that are not captured by gene expression alone.

This UROP project will explore how deep-learning-based image features can be extracted and integrated with spatial gene expression data using R-based tools within the Giotto Suite ecosystem (www.giottosuite.com). The student will help evaluate different image tiling strategies (large vs. small regions of tissue), test pretrained deep-learning models, and explore how image-derived features relate to molecular states in cancer tissues.

The goal is to develop well-documented, reproducible workflows that make advanced image analysis more accessible to researchers working in R. This is an excellent opportunity to gain experience in computational biology, machine learning applications in biomedicine, and open-source software development. In addition, you will be able to work with large-scale spatial datasets generated from primary tissues of patients at BMC.

We specifically welcome students with experience or coursework in programming (R or Python), data science, statistics, or biology. Enthusiasm for learning and translational research is essential.

Students will receive mentorship from Drs. Ruben Dries and Jiaji George Chen, and other members of the Dries Lab, with opportunities to interact with both computational and wet-lab researchers.

A commitment of at least 10 hours per week is expected, with flexibility based on the student’s academic schedule.

To apply, please send your resume, a short statement of interest, and information about previous experience with R programming, deep-learning fundamentals, and spatial omics, to:

Ruben Dries (rdries@bu.edu) & George Chen (jiajic@bu.edu)

Back to On-Campus Opportunities